The first one is the Republican party. The Republican party has put forward a strong presence for a long time now. Indeed, it is the puffery of their bluster that forms the basis of their self image. I think they do this for several reasons (many described by you over the years): first, the base of their party is men and men like to have the feeling of voting for a winner much more than women (http://www.scientificamerican.com/article.cfm?id=vote-election-testosterone). Second, I am sure some undecideds want to vote for the winner just like I like to pick a winner at the end of the NHL season--better to be part of a winning team. In each of these, the Republicans use this facade of strength to accomplish their electoral goals. For example, the Republicans have spent decades complaining about a liberal media bias and have pushed the press to the point of constant he said/she said articles. Moreover, with their ability to project a single message so clearly at all levels, the media is pushed to repeat their message. Example: Romney has momentum. When Nate Silver says repeatedly, with data, that this is wrong, it knocks off a key part of their strategy. He was killing all three Republican needs at once: Romney wasn't winning, the media was harder to push around, and false bravado makes you look weak. Indeed, I worried about Karl Rove's math in 2006 before the midterms, but this time I smiled and recognized the bully trying to look tough.
The second group that I think doesn't want Silver to be right are the horse race pundits. Just think of the CNN folks after the debates in 2008 and 2012. They would say what they think and then the data would come in and they would have to change what they said. They looked dumb. Data couldn't be thrown away. The pushy pundit couldn't walk over the weak one. Data equalizes everyone. Well, why pay Joe Scarborough a lot of money when his analysis is unreliable? He has a financial interest in proving his visceral understanding of the race is more cogent than Nate Silver's statistical one. That is why the bet is so brilliant. It is a material representation of what is happening. As Nate wins, Joe loses. This election may be best remembered as the occasion when Money Ball came to politics and won. When that happens, who do you want on your news station or news paper? The people who are right, of course. In fact, I would wager that Margaret Sullivan is missing the larger picture of her own demise in media. A public editor is needed most when two opinions are at odds. When solid data is available, one side can argue the world is flat, but you don't need a public editor mediator to solve that debate, you need a map.
In short, better information available from the variety of statistical websites is better for humanity, but worse for the pundit's mortgages, and thus the disconnect.
I'll have to reserve judgment on the Republican party in this analysis. But I think there is very significant status anxiety on the part of what JB identifies as horse race journalists. Not that there's anything wrong with being a horse race journalist. I consider myself one. Guilty. I love the horse race. Show me any honest political junkie who's not. But the real coin of the realm of talking or writing about politics should be insight, not saying I just talked to this person who doesn't know anything either and so I must know more than you do.